The Application - AgAdvantageTM was a massive enterprise processing and account management application for agricultural insurance policies. It grew from a small legacy app with limited functionality to essentially process and maintain the entire company's business. The developer driven application had grown without the end user in mind and had no attention paid to the information architecture.
Pain-point for Partners - The application had become a wedge between partnering agents and the company. They loved working with our people, but loathed using the application to process and administer client policies. The system was slow and it took multiple, convoluted clicks to get simple information. Agents would often have to tell their customers that they would "have to get back to them" when on the phone. The competition often had newer, more modern systems that were easier to use. At the end of the day, the partnering agencies were the ones who brought us business. We had to focus on their needs or risk losing more business.
Pain-point for the Business - The application was role and permission based, so the business was working from the same pages as the agents. They faced the same issues: slow response, poor system performance and slow navigation. This created a slew of trickle-down effects on business productivity and worked to drive a wedge between the business and IT. This created an adversarial relationship with one always blaming the other.
User Research and Interviews - The first thing we needed to do was speak directly with the users, both business and partnering agencies. We needed to understand them, their roles, their issues, constraints and daily activities. My team and I planned and executed several daily calls with business users and agency users. I was able to convince the business that we could do a lot to mend relationships with agencies by inviting them to the office for face to face interviews, user testing and to share roadmap items. This had a great impact and became a regular event after the success of the first two.
Educating the Business - Momentum was on our side, but there was still some push-back from the business. The business analyst manager and development manager and I decided to travel to headquarters and hold a design thinking ideation session with the business heads. We would not only educate them on the importance of user-centered design, but also have them help in creating the solution. Ultimately, they helped us produce personas, empathy maps and prioritize user needs, based on their keen insights. Many of them were former agents and understood both sides of the problems intimately.
Metrics and Analytics - The next thing we needed to do was get a baseline of where we were from a quantitative viewpoint. We gained initial metrics from user testing like "time on task" as a start. We also added Google Analytics to the application to start getting insight on what users were doing and where they were spending most of their time. We started to see patterns emerge that helped further lead us in the direction of a solution.
A Dashboard and Global Search - Working with the business leadership, coupled with our user interviews and research, helped to close gaps and paint a good picture of how to solve the problem. We hypothesized that a role-based dashboard and global search would solve multiple pain-points at once. Not only would we help them prioritize what policies needed managing, premiums due and the status of claims for customer relations, but we also gave them a quick and direct pathway to the information they needed. Users would now be able to get to accounts, policies or claims from anywhere in the system at any time.
Wireframing Process Flow - My team and I outlined all of the features and capabilities we believed to be necessary for users to be efficient and productive from the time they logged in, based on the input from agents and the business. We decided to focus on Agents only for the initial rollout, as they were a greater priority from the company leadership's perspective. Plus, it was an easy way for us to iron out the bugs with one user cohort before applying the solution to others.
Prototyping and Low-fi Testing - Once we had the elements and process hammered out, we designed high-fidelity, quasi-functional prototypes with HTML, CSS and JQuery to put in front of agent users bot remotely and during agency visits. We gave them scenarios to work with to see if they would act the way we predicted. We discovered a few issues that we had to address right away. Otherwise the tests and follow-up interviews confirmed our assumptions. With our issues addressed, we knew we could move forward in development with confidence that we were headed in the right direction.
Learn and Iterate - After the initial launch, we discovered new, minor issues we needed to address. They were easily changed and added to the next sprint. We tested regularly for some time after the launch to make sure the changes and additions met user needs and expectations.
UX Was Foreign to the Business - The largest challenge my team and I faced was convincing the business to prioritize user-centered design. Each department head had their own issues that they wanted IT to address before the agents. It took bold leadership to over-ride those voices. Leading a design thinking workshop with the business executives was not easy, as I was flying solo and that is not a great way to lead a session of that size. In the end, it helped get them on board and gave them a better understanding of why it was important to support user-centered design.
Issues with Remote Testing - Remote testing can work well, though not optimal. New applications make it easier, but issues can still arise. We ironed out a lot of kinks along the way and improved the process over time.
Assumption Mis-steps - There were a few key items on the dashboard and global search that we overlooked. Luckily, these items were caught before development in testng with quasi-functional prototypes. This saved a lot of time, money and resources in the end.
"Game Changer" - The results were profound. Our facebook page lit up with comments, customer success managers and sales managers all received calls and emails raving about the new dashboard and global search. A common comment we all loved was that agents said it was a "game changer" for them and their workflows.
Net Promoter Score - Our users were delighted overall and our improvement in net promoter scores affirmed what we were hearing verbally and via email.
Improved Relationships - The new changes and working with agencies went a long way in improving relationships across the board. The business and IT were communicating more and reacting less. Even the business received immediate improvements to their navigation. What once took them several minutes to complete, now took seconds. I could not have been happier with my team or the results.